1906. No. 3. AMUNDSEN'S OCEANOGRAPHIC OBSERVATIONS IN 1901. 9 
tight, and the evaporation has in fact been insignificant, except in a very 
few cases. 
The water-samples were examined by titration (Mohr) at my labora- 
tory in the University of Christiania, as soon as possible after they were 
received. The titrations were carefully made by Mr. Ingolf Leivestad; 
they were frequently controled by my assistant Mr. Jakob Schetelig, 
whose determinations with the Hydrometer of Total Immersion agreed on 
the whole well with Leivestad’s results. Mr. Leivestad controled every ten 
titrations by means of Standard Water, determined by Mr. Schetelig with 
Hydrometer of Total Immersion. 
Mr. Schetelig gives the following information about the Standard 
Water and Leivestads’ determinations: 
“The Standard-Waters were obtained by mixing several samples of 
sea-water in a big glass-bottle, which was afterwards filled into green 
selters-bottles with patent india-rubber stoppers; 8 or Io bottles were in 
this manner obtained of each water. Standard Waters Nos. I, II, and II 
were used by Leivestad for about 3000 titrations. The values for the 
salinity of the Standard Waters, were exclusively based on determinations 
of the specific gravity by means of the Hydrometer of Total Immersion. 
The following final values were obtained as the means of a series of 
determinations. 
99 S °/oo Cl. °/60 
Sept. 9, 1901 Standard-water No. I . . . 2771 34°49 19°09 
Nov. II, 1901 — å i” AE 28:00 34°85 19°29 
Jan. I, 1902 = SAN crises. 028702 34°87 19°30 
“The accuracy of the value of o, is probably inside a limit of + 0'005 
(i. e. inside + 0'000005 of specific gravity). 
“Standard-Water No. I has unfortunately a lower salinity than a 
Standard-Water ought to have, and this water was used for the titration 
of all samples of deep water from Amundsen’s Stations. The same water 
(No. I) was also used for the titration of Makaroff’s samples (see below). 
“In order to use Mr. Knudsen’s Titration-Tables a series of aid-tables 
had to be computed. 
“If by the use of this Standard-Water No. I, with the low salinity, 
any error at all has been introduced into the determinations, this error 
must be constant for all of them, and may be computed.” 
“By comparing the results of the titrations I have, however, come to 
the conclusion, that there cannot be any great error. The values ob- 
tained by titration and by the Hydrometer of Total Immersion agree very 
